DRVISION wins NIMH SBIR grant to develop the Intelligent Connectomics Analysis platform within Aivia
Updated: Aug 7, 2019
Full title: DRVISION wins up to 3.27 million USD NIMH SBIR grant to develop the Intelligent Connectomics Analysis platform within Aivia.
Bellevue (WA), 1st of July 2019 - DRVISION Technologies, a pioneer in computer vision and the world’s leading AI microscopy software company has been awarded up to 3.27 million USD from the National Institute of Mental Health (NIMH) to develop the Intelligent Connectomics Analysis (ICA) platform. The R&D funded by this Small Business Innovation Research Fast-Track grant (# R44 MH121167) will further augment the machine learning enabled solutions currently available in Aivia and Aivia Cloud.
ICA will greatly automate the labor-intensive task of manual neuron reconstruction, synapse detection and circuit inference based on Electron Microscopy (EM) and Correlative Electron and Light Microscopy (CLEM) images by creating and deploying innovative AI assisted tools. ICA tools will handle multi-terabyte data sets, be usable by neuroscientists with no machine learning expertise and will cover all major aspects of the connectomics analysis pipeline: ground truth creation; training of deep learning models for neuron boundary detection; applying said models, neuron tracing, synapse detection, circuit determination, editing of neurons, synapses and circuits; interactive visualization of raw and processed data; and downstream data driven analysis such as functional inference from structure and phenotypic discoveries. Moreover, ICA will pilot the use of “bi-directional active learning”, a groundbreaking AI powered method which guides users to areas that likely need human corrections, makes use of the updated annotations to further improve the initial AI module and incrementally offers additional high confidence neuron reconstructions until the full 3D dataset is mapped. ICA tools will be commercially supported and integrated into Aivia to accelerate the rate of knowledge creation in the field of neurosciences.
DRVISION will work closely with Prof. Rachel Wong from the University of Washington in addition to 10 other collaborating sites from across the USA (University of Washington, University of California San Francisco, Northwestern University, University of British Columbia, Washington University, University of Oregon, Stanford University / Howard Hughes Medical Institute and the University of Chicago). The collaborating sites will provide test data sets, detailed user feedback and validate ICA’s tools using different samples from different regions of the nervous system and across key model systems such as mouse, zebrafish, fly and worm.